Assimilating goes-r products into an ensemble based numerical weather prediction model to enhance short-term probabilistic forecasts of tornado producing storms we propose to extend a GSI-based enkf…

This federal contract award to the University of Oklahoma (OU) is for research to assimilate data from the upcoming GOES-R satellite into an ensemble-based numerical weather prediction model. The goal is to enhance short-term probabilistic forecasts of tornado-producing storms. The $346,734.75 grant, awarded by NASA, will allow OU to develop tools to reliably assimilate GOES-R satellite data, including radiances, cloud property retrievals, and atmospheric motion vectors, along with traditional radar data. This research supports the Warn-on-Forecast initiative to improve short-term prediction and warning lead times for severe weather events like tornadoes. OU, a renowned public research university with expertise in atmospheric sciences and meteorology, will leverage its academic and research capabilities to advance this scientific and operational objective without any set-aside designation.

Assimilating goes-r products into an ensemble based numerical weather prediction model to enhance short-term probabilistic forecasts of tornado producing storms we propose to extend a GSI-based enkf data assimilation system to generate probabilistic forecasts of rotating supercell storms that produce tornadoes. The upcoming launch of the goes-r satellite with the advanced baseline imager (ABI) will provide visible and infrared radiances, cloud property retrievals, and atmospheric motion vector (AMV) winds at spatial and temporal resolutions comparable with surface based radar data allowing for the first time a true comparison of the effectiveness of each in a high-resolution numerical weather prediction (NWP) environment. Accomplishing this task will require development of the tools necessary to assimilate goes-r data on these scales in a reliable manner in concert with traditional high resolution radar reflectivity and radial velocity data. Satellite radiances will be used to improve the characterization of atmospheric water vapor content in clear sky regions whereas cloud property retrievals such as cloud water path (CWP) and cloud top temperature (CTT) will be used to characterize cloudy regions initially, with a focus on direct assimilation of cloudy radiances occurring during the latter stages of the proposed research. Amvs will also be used to characterize the wind fields where gaps in doppler radial velocity observations may exist. Finally, radar reflectivity will be the primary source for observations of deep convection and heavy precipitation. By combining these various data types, IT should be possible to improve both the characterization of the near-storm environment and the storms themselves in ways not possible using either satellite or radar data alone. This represents one of the major goals of the warn-on-forecast (WOF) initiative whose mission is to improve the short-term prediction of severe weather events and warning lead times for tornadoes. The following proposal describes in detail our plan to further advance this goal.